278 
SCOTT'S LAST EXPEDITION [February 
rising and falling with the swell. We reached this across 
an ice island, surging about in a large pool. In spite of 
all this movement, no more of the ice moved north as 
far as we could judge. 
On the evening of February ist I held a council. 
Captain Scott's instructions read : ^ I am of the opinion 
that the retreat should not be commenced until the 
bays have refrozen, probably towards the end of March. 
An attempt to retreat overland might involve you in 
difRculties — whereas you could build a stone hut, pro- 
vision it with seal meat and remain in safety in any 
convenient station on the coast.' 
However, he gave me permission to begin the 
retreat in February if we were not relieved in January, 
and I began to prepare for this event, for I felt sure we 
could traverse the piedmont glacier. 
Cracks seemed to be spreading on the sea ice even 
while one was watching it. The surging ice-blocks in 
the tide-crack, now twenty feet wide, rose several feet. 
Now and again a huge shock, as of a big rock bumping 
on another, announced a new crack, while a constant 
roar, like that of a distant lion, announced the periods 
of maximum of the swell rolling in from twenty miles 
away. 
On February 3 Debenham, Gran, and I climbed the 
glacier slope behind our camp to prospect for a path. 
We roped up and proceeded about three miles southward, 
keeping well behind the crevasses. These are numerous 
on the steep seaward slope, but we met with none on 
the fairly level ground, though we could see them just 
